Download The Psychology of Global Warming - American Meteorological Society

The Psychology
of Global Warming
Improving the Fit between the Science and the Message
by
Ben R. Newell and Andrew J. Pitman
Insights from the psychology of judgment and decision making might help the climate
community communicate global warming science to an often skeptical public.
C
onsider the following question: If a bat and a ball
cost $1.10 in total and the bat costs $1 more than
the ball, how much does the ball cost? If you are
like many people, your immediate answer would be
10¢ (Kahneman and Frederick 2002). You would be
wrong. Think a little and you will see why. This is a
simple question and the answer is easy, but people
still often get it wrong. Contrast the simplicity of
this question with the complexity of global warming.
Is there any hope of communicating such complex
issues if many people fall foul on a simple mental
arithmetic problem? Yes. These simple examples
and the resulting errors can illuminate how we make
decisions. Knowledge of these mechanisms can help
AFFILIATIONS: Newell—School of Psychology, University of
New South Wales, Sydney, Australia; Pitman —Climate Change
Research Centre, University of New South Wales, Sydney,
Australia
CORRESPONDING AUTHOR: Ben R. Newell, School of Psychology, University of New South Wales, Sydney, 2052 NSW,
Australia
E-mail: [email protected]
The abstract for this article can be found in this issue, following the
table of contents.
DOI:10.1175/2010BAMS2957.1
In final form 9 March 2010
© 2010 American Meteorological Society
AMERICAN METEOROLOGICAL SOCIETY
us understand how to communicate climate science
to fit with the way humans process information. Our
aim in this paper is to provide some suggestions for
improving this fit.
Evidence for warming. The idea that Earth is warming
partly because of the emissions of greenhouse gases
to the atmosphere is one of the most certain concepts in natural science. The idea that greenhouse
gases increase radiative forcing is an old idea that
has withstood a variety of analyses to emerge intact
(an accessible history is available on the Web site of
the American Institute of Physics at www.aip.org/
history/climate/co2.htm). The peer-reviewed papers
that provide the evidence that human-induced emissions of greenhouse gases over the twentieth century
have led to increases in temperature and changes in
rainfall, wind, humidity, sea level, ocean acidity, snow
cover, etc. have been assessed rigorously through the
Intergovernmental Panel on Climate Change (IPCC)
in a series of reports. No serious academic body, significant institution, or national government doubts
the basic science (e.g., Somers 2009).
Despite this near total lack of evidence to the
contrary, a significant portion of the public, journalists, and politicians emphasize their serious doubts
about the science of global warming. This skepticism has increased as the level of scientific certainty
in global warming has increased. For example, in a
AUGUST 2010
| 1003
recent survey of U.S. residents, only 57% said that
they believed global warming was happening, down
from 71% in 2008 (Leiserowitz et al. 2010). Why has
this “disconnect” (Mariconti 2009) between the science and the public/media perception of the problem
occurred?
Many readers of the Bulletin of the American
Meteorological Society (BAMS) routinely give talks
based on the IPCC science. Clearly, the message
that this science is robust and leads to a conclusion
that we must act aggressively to reduce emissions is
either not being heard, being dismissed, or not being
acted on. This failure to get a clear message through
is presumably one of the reasons for this disconnect.
Thus, some effort to refine and hone the way in which
climate scientists communicate their findings to the
wider community appears warranted.
Interpreting evidence. How humans interpret evidence,
how they react to evidence, and how they form views
based on evidence are not related merely to the quality of the evidence. Psychologists, especially those
with an interest in judgment and decision making,
explore precisely these issues and can therefore
provide insights into the impediments confronting
the communicators of climate science. The role that
psychology can and should play in the climate change
debate was heralded almost 30 yr ago (Fischhoff 1981)
and since then there has been an increasing amount
of research at the intersection of psychology and
climate science, especially in the last decade (e.g.,
Budescu et al. 2009; Fischhoff 2007; Fischhoff and
Furby 1983; Hardisty and Weber 2009; Leiserowitz
2006; Moser and Dilling 2004; Nicholls 1999; Sterman 2008; Stern 1992; Weber 2006). Our aim in this
paper is not to provide a detailed synthesis of this
literature: many relevant topics have been covered in
much greater depth by other authors (e.g., Morgan
et al. 2002); more modestly, we want to make BAMS
readers aware of this literature, highlight what we
find to be thought-provoking and relevant insights,
and offer some brief suggestions for improving communication of the complexities of climate science. For
more specific communication strategies, interested
readers should consult some of the excellent resources
listed in the “Further Reading” sidebar.
Insights from psychology. We have divided the paper
into four sections, each of which focuses on a different “class” of psychological phenomena: sampling,
framing, comprehension, and the process and perception of consensus building. The presentation of
the psychological phenomena is preceded, in each
1004 |
AUGUST 2010
section, by a statement of a problem from the climate
perspective along with reasons why the problem
often leads to confusion on the part of the (lay)
audience. The relevant psychological phenomena
along with suggestions for how to “tackle” them are
then presented primarily in the four main tables of
the article. The accompanying text in each section
provides further explanation of the phenomena listed
in the tables and references to relevant articles in the
psychology literature.
Sampling Issues : What samples
of evidence do people use when
making judgments? By a “sample,” we mean
the subset of information that a person uses to draw
an inference or conclusion. We identify two key
sample problems in global warming science: “weather
versus climate” and “is it warming?” We will briefly
summarize these problems before presenting some
relevant psychological phenomena.
Climate phenomenon: Weather versus climate. No
BAMS reader would confuse weather and climate,
but the media and the general public routinely do.
Who has not been confronted by “it was cold this
spring, that global warming thing must be a myth”?
You can replace “spring” with a single day, week,
month, year, or decade. Or someone saying, “I remember it being much hotter when I was a child.”
The IPCC addressed this issue in a frequently asked
question in the Fourth Assessment Report (FAQ 1.2)
and explained the nature of weather versus climate,
the role of chaos, and the difference between a deterministic weather prediction and a statistical climate
projection. However, the explanation is complex to
the journalist, the policy maker, and the general
public, despite every effort to use simple language.
Understanding the explanation requires an appreciation of how the climate system works, through
time, on a variety of spatial scales. Understanding
time scales, the interactions of forcing and natural
variability, evolution of various components of the
climate on different time scales, and the projection of
these onto changes in various systems is basic climate
science. It is not, however, what the vast majority of
people think about.
Climate phenomenon: Is it warming? Earth is warming
and it is warming in large part because of the human
emissions of greenhouse gases. The IPCC Fourth
Assessment Report (specifically Working Group 1)
provides a convincing defense of this statement. Of
course, not every day is warmer than yesterday, not
every year is warmer than last year, and not every
decade has to be warmer than the last decade. Some
recent evidence points to the probability of some
decades through the twenty-first-century cooling
relative to the previous decade, whereas the overall
trend through this century would be one of warming
(Easterling and Wehner 2009). There will inevitably
be some cold months, some deep snow falls, or some
very cold days through the twenty-first century. Indeed, an intensification of the hydrological cycle or
stronger storm tracks may increase the likelihood of
some extreme cold events in the future. Global warming is, of course, about trends and averages at large
spatial scales, and the averages are not commonly
grossly affected by small changes in the frequency
of rare events.
Psychological phenomena influencing samples of evidence.
Table 1 identifies four psychological phenomena relevant to the sampling issues described earlier. With
regard to the weather for climate problem, a relevant
phenomenon is that of “attribute substitution,”
whereby people tend to substitute difficult questions
with ones that they find easier to answer (Kahneman
and Frederick 2002). The bat-and-ball problem
introduced earlier is a good example. The amount
of $1.10 is easily decomposed into $1 and 10¢; 10¢
sounds about the right price for a ball, so we give that
answer, even though a little more thinking tells us it
is wrong.1 Thus, when asked to think about changes
in the climate (a complex spatial and temporal conThe correct answer is 5¢.
1
Further Reading
T
his is an annotated list of reading we recommend. We begin with something roughly equivalent to the frequently asked questions from the IPCC. We move from this through accessible “popular science” books written by academics all of which are
consistent with the psychological literature. Finally, we recommend more specialized texts for the climate scientist who wants
to develop a deeper understanding of the psychology of judgment and decision making
The Psychology of Climate Change Communication: A Guide for Scientists, Journalists, Educators and the Interested Public by The
Center for Research on Environmental Decisions (downloadable for free online at http://cred.columbia.edu/guide).
This guide was prepared by a group of highly respected psychologists at the Center for Research on Environmental
Decisions (based at Columbia University). It is a very readable overview of a wide variety of peer-reviewed work examining several facets of the psychology of climate change communication. An invaluable resource for climate scientists wishing to improve public presentations of their science.
Reckoning with Risk: Learning to Live with Uncertainty (U.S. title: Calculated Risks) by Gerd Gigerenzer. Published by Penguin.
Gigerenzer is the Director of the Adaptive Behavior and Cognition Group at the Max Planck Institute for Human
Development in Berlin. He has published widely on risk communication and this book brings together much of his work
on how to understand and present statistical information.
Nudge: Improving Decisions about Health, Wealth and Happiness by Richard Thaler and Cass Sunstein. Published by Penguin.
Thaler, an economist, and Sunstein, a lawyer, are both based at the University of Chicago. They have both published
extensively on the intersection of psychology, economics, and law. This engaging book is an investigation into how an
appropriate “choice architecture” can improve our decision making.
Risk and Reason by Cass Sunstein. Published by Cambridge University Press.
This is an examination of how reasoning about risk through a cost-benefit analysis can improve risk regulation. The book
includes particular focus on environmental risks (e.g., global warming).
Risk communication: A Mental Models Approach by M. G. Morgan, B. Fischhoff, A. Bostrom, and C. J. Atman. Published by
Cambridge University Press.
A highly detailed yet very accessible “field guide” to the use of the mental models approach to risk communication. It
includes in depth discussion of global warming and climate change. The authors are experts in public policy, psychology,
and engineering.
Straight Choices: The Psychology of Decision Making by Ben R. Newell, David A. Lagnado, and David R. Shanks. Published by
Psychology Press.
An up-to-date and accessible text book-style introduction to many of the key findings in the psychology of judgment and
decision making. The authors are all psychologists and have published in many areas of cognitive psychology including
papers on learning, memory, causal and probabilistic reasoning, and decision making.
AMERICAN METEOROLOGICAL SOCIETY
AUGUST 2010
| 1005
Table 1. Sampling issues: psychological factors that can influence the samples of evidence people use when
making judgments.
Psychological phenomenon
Illustration
Suggestion for tackling phenomenon
Attribute substitution
Judgments about attributes that are
inaccessible or not easily understood,
such as “climate,” may be substituted by
more readily accessible attributes, such
as “weather.”
Always define terms clearly, even those that
are commonplace within the discipline. Invite
your audience to reflect carefully on their
responses to questions to ensure that the
question they answer is the same as the one
being asked.
Recency of events in memory
A judgment about the reality of global
warming may be more strongly influenced
by recent weather events (a cold snap
or intense summer heat) because recent
events in memory are more salient than
those experienced longer ago.
Invite your audience to think about the samples
that they are using to draw conclusions. For
example, ask them to place recent salient
events (e.g., a cold snap) into a wider context,
such as the last 10 yr or their whole lives.
Insensitivity to bias in external
samples of information
Judgments about the “balance of evidence”
in relation to climate science might be
unduly influenced by disproportionate
media coverage given to those who
dispute the basic science.
Remind your audience of the distinction
between the debate within the climate science community and the debate as presented
via the media. Encourage them not to confuse
these separate debates when evaluating the
science.
Anchoring on irrelevant
information
In conditions of uncertainty, estimations
of magnitude (e.g., the concentration of
atmospheric CO2) can be influenced by
the provision of (irrelevant) reference
points (e.g., is it below 1000 ppmv vs
above 100 ppmv?)
When presenting data identify samples
clearly, provide justification for the size of the
sample (e.g., a time span of 10 or 100 yr) and
choose reference points carefully.
cept), people might instead think about changes in
the weather (a concept familiar to everyone).
This confusion can be compounded by the second phenomenon, recency, the finding that events
that have occurred more recently are more salient in
memory and thus tend to have a disproportionate influence on our judgments (Weber 2006). Thus, when
people think about whether the planet is warming,
events such as the extreme cold weather of the 2009/10
winter in parts of the Northern Hemisphere may inappropriately affect judgments simply because those
events are current and memorable.
Furthermore, the third phenomenon, biases in the
external samples of information, can affect memory
and judgment processes. For example, if the public
read or hear opinions from climate change skeptics
about 50% of the time, then this could lead to a bias in
the perception of the balance of evidence in the minds
of the public (i.e., that the science is only about 50%
certain; Moser and Dilling 2004). Failure to accommodate for this bias in the input sample necessarily
leads to erroneous judgments about the likelihood of
future outcomes (Fiedler 2000).
1006 |
AUGUST 2010
The fourth sampling phenomenon identified in
Table 1 is that of anchoring. This is the highly robust
finding that the reference point or “anchor value”
one uses in describing samples can affect peoples’
judgments and magnitude estimations (Tversky and
Kahneman 1974). For example, if you ask an audience
to first state whether the current concentration of
CO2 in the atmosphere is greater than 1000 ppmv and
then ask for their own estimate of the concentration,
they are likely to give a much higher estimate than
an audience who are first asked if the concentration
is greater than 100 ppmv. When people are uncertain
about a true value, given values such as 1000 and 100
are often interpreted as appropriate anchors, which
are then insufficiently adjusted away from when
making an estimate.
The brief suggestions for tackling these four phenomena, listed in Table 1, all follow a common theme:
think carefully about the samples of evidence that
you use in presentations, and encourage audiences
to reflect on the samples that they are using to draw
their conclusions. Sometimes a simple instruction to
your audience to consider the opposite conclusion—
that is, engage in some thinking about why an initial
hypothesis or preconception might be wrong—can
counteract sampling and anchoring biases. By
directing their attention to contrary evidence that
might not otherwise be considered, people can
expand their samples of evidence, thereby making
them more representative (Larrick 2004; Strack and
Mussweiler 1997).
Framing issues: How does the way
evidence is presented affect perception and reaction ? Framing is a
term used to describe how information is presented
to convey a particular message or to produce a desired response. In this section, we first describe why
choosing the appropriate frame for the science of
global warming is as difficult as it is important; then,
in Table 2, we present four psychological phenomena
relevant to framing.
Climate phenomenon: CO2—Odorless, colorless, and
rare. In the Garnaut Review (an Australian version
of the Stern Review), global warming was described
as a “diabolical” problem. It is worth ref lecting
on the reasons for its status as diabolical. Carbon
dioxide is a colorless, odorless gas. If it was opaque or
smelled like methane, we would never have allowed
it to increase to around 390 ppmv, but like most
chemicals it becomes a pollutant when increased
beyond its normal concentration. It is a natural
gas, part of the natural carbon cycle, and essential
to plant life. There is no single significant source;
it is simply the by-product of most human activities. We measure the impact of increasing CO2 in
terms of radiative forcing. The increase in radiative
forcing to date has been about 1.6 W m−2 because
of all forcing and an increase of about 2.64 W m−2
from the long-lived greenhouse gases. This small
number sounds entirely trivial, but 2.64 W m−2 is
about 8,322,502,000 J m−2 yr−1.
Carbon dioxide is, of course, very rare; it makes
up about 0.0384% of the atmosphere by volume.
Collapsed to a single layer, it is about 8 m deep. Of
course, the amount of something need not be proportional to the impact. For example, cyanide at a level of
concentration 100 times lower than the concentration
of atmospheric CO2 in the blood (0.0003%) is fatal.
A further problem is that it has taken a century of
human emissions to register a problem, and we tend
to talk about major problems occurring in, for example, 2050, which is almost an infinitely long period
into the future for most human decision making. It
also takes decades for the impact of a given level of
AMERICAN METEOROLOGICAL SOCIETY
atmospheric CO2 to be realized in temperature, sea
level rise, and ice sheet melt.
Finally, if humans dramatically cut emissions,
it would take decades before we knew if we had
prevented dangerous warming and centuries before
natural levels would be realized. Moreover, like the
Y2K bug, no consequences could easily be interpreted
as “there was never a problem” rather than “we acted
to avoid a problem.” If you wanted to create a diabolical problem, it is hard to imagine doing it better than
releasing CO2 into the atmosphere.
Psychological phenomena relevant to framing. The first
phenomenon described in Table 2 makes the simple
point that the numbers one uses to convey the statistics of global warming can have huge impact on
interpretation of the severity of the problem. As we
noted in the previous section, a change in the units
used to describe radiative forcing can turn a small
unimportant sounding number into a very large,
more worrying one. Alternate ways of expressing
uncertainty can also lead to dramatic differences in
interpretation. Several lines of research suggest that
numerical information expressed in a frequency format (e.g., 1 out of 100) is more easily interpreted and
reasoned with and has a greater influence on judgments than identical information presented as percentages (1%) or probabilities (0.01; e.g., Gigerenzer
and Hoffrage 1995; Newell et al. 2008; Slovic et al.
2000; Yamagishi 1997).
The second phenomenon in Table 2 provides
insight into a reason why frequency formats (i.e., 1
out of 100) have this effect. Information processing
is influenced not only by “cold” cognitive processing
but also by “hot” emotional or affective processing
(Loewenstein et al. 2001). Global warming is a highly
emotive issue, but attaching imagery or affective
valence to the particular effects of CO2 is difficult
because it is colorless, odorless, and slow acting.
Framing outcomes in terms of numbers like 20 out
of 100 (rather than 0.2 or 20%) can engage affective
processing because they feel more concrete to individuals than probabilities and percentages (Slovic
et al. 2000).
Making outcomes feel more concrete can also
help in addressing the third phenomenon in Table 2,
the tendency for people to discount the importance
of future events relative to those occurring now
(Hardisty and Weber 2009; Trope and Lieberman
2003; Weber 2006). One explanation for this discounting of the future is that people tend to construe distant
and near-future events very differently. Distant future
events (such as the prospect of sea level rises in 50 yr)
AUGUST 2010
| 1007
Table 2: Framing issues: psychological factors that can influence the perception of and reactions to
evidence.
Psychological phenomenon
Illustration
Suggestion for tackling phenomenon
Psychological nonequivalence
of mathematically equivalent
information
A chance of 1 in 1000 and 0.1% are
mathematically but not psychologically
equivalent. Representing the chances of
occurrence as frequencies rather than
percentages or probabilities can improve
the perception and understanding of a
complex problem (e.g., the chance that
CO2 will increase above “safe” levels).
Try to choose one type of numerical format
in your presentations (e.g., a frequency
format; 1 in 1000) and be consistent in its
use throughout. When possible, use simple
graphs (e.g., pie charts) to convey numerical
information.
Influence of affective processing
of information
Information processing does not occur
in an emotional vacuum; processing of
the affective content of information or
the emotional reactions that information
evokes contributes strongly to perception
and understanding of evidence (e.g., the
effect of increased CO2)
Use vivid images of global warming (e.g.,
shrinking glaciers, melting ice sheets) to
engage emotional processing, but do so
judiciously to avoid emotional numbing or a
“despair” response.
Discounting the importance of
future events
Judgments about the importance of future
events (e.g., perceived impacts of global
warming) tend to be discounted relative
to events happening in the present.
To mitigate discounting, try to use specific
and where possible concrete examples of
distant future outcomes (e.g., the appearance of your audience’s local environment
in 30 years time or the air quality that their
children might face in 2050).
The differential impact of losses
and gains
Losses of a given magnitude inflict more
psychological “pain” than gains of the
same magnitude satisfy. Thus, people tend
to avoid outcomes framed as losses more
than seek those framed as gains.
Capitalize on “loss aversion” by explaining
how actions to mitigate global warming
(e.g., insulating one’s home) will lead to the
avoidance of large losses (i.e., higher energy
bills).
are represented abstractly in terms of general features
and the “essence” of an event. In contrast, near-future
events (such as the prospect of a river near your
house flooding its banks tomorrow) are represented
more concretely in terms of the specific contextual
and incidental details (e.g., the effect a flood will
have on the carpet in your living room; Trope and
Lieberman 2003; Weber 2006). Encouraging people
to think about the possible specific impacts of future
events in the context of where they live and how these
events might affect their daily routines may have the
dual advantage of engaging affective processing and
counteracting the tendency to discount the future.
One note of caution: research suggests that people
can become “numbed” by overuse of emotional
appeals and that they can only worry about a limited
set of issues (a “finite pool of worry”; Linville and
Fischer 1991; Weber 2006). Thus, although vivid images and concrete outcomes are important when presenting the science, one should use them judiciously.
Overuse may have the unintended consequence of
1008 |
AUGUST 2010
leaving the audience overwhelmed and thus unwilling
to take any action on what they perceive as a fait
accompli (Moser and Dilling 2004).
The final phenomenon in Table 2 refers to the
well-established finding that losses and gains have
a very different psychological impact: the pleasure
associated with receiving $500 is less than the “pain”
felt when one loses the same amount (Tversky and
Kahneman 1981, 1992). This asymmetry in the effect
of gains and losses leads people to be “loss averse,” a
tendency to be more averse to losses than be attracted
by corresponding gains. Recent research indicates
that some environmental outcomes are treated similarly to financial ones (Hardisty and Weber 2009);
thus, when describing actions to mitigate global
warming, messages should focus on the potential to
avoid large losses (e.g., high fuel or heating bills) than
the corresponding gains (e.g., the savings accrued
over time by installing solar hot water).
The suggestions for tackling these framing
phenomena are, like those for Table 1, simple to
summarize. To begin with, do not assume that your
audience will interpret numerical information and
figures as you do (you are unlikely to be speaking
to an audience that is as familiar with quantitative
relationships as you). Thus, numerical information
should be conveyed when possible using an easy-tounderstand format (frequency formats have many
advantages) combined with straightforward graphs
[e.g., pie charts rather than probability density functions (PDFs)] with a minimum of potentially opaque
abbreviations such as ppmv or 10 6. Concrete, specific, and vivid outcomes of global warming should
be emphasized and combined with examples that
emphasize the avoidance of future losses rather than
the receipt of future gains. However, this needs to be
done cautiously to avoid emotional numbing.
Comprehending the problem and
the solution: How the construction of mental models affects the
conceptual understanding of
global warming. “Mental model” is the term
used to describe the collection of partial knowledge
and beliefs that people “build” to help understand and
make decisions about a given problem or phenomenon (e.g., global warming; cf. Morgan et al. 2002). In
this section, we discuss the aspects of global warming
that make forming appropriate mental models difficult and make some suggestions for how audiences
could be assisted in building a scientifically informed
mental model.
Climate phenomenon: The problem and the solution.
The solution to global warming is an economic,
engineering, legal, social, and political challenge
that is perhaps unprecedented in recent history. It
“merely” requires a transition from a carbon-based
economy to a non-carbon-based economy. This need
is interweaved with a highly complex debate about the
level of risk associated with global warming and the
scale of solutions required to resolve the problem.
At the heart of the solutions debate is the issue
of targets. These are commonly thought about in
terms of the amount of warming that is “safe” (e.g.,
2°C; Pachauri and Reisinger 2007). We do not know
what is safe because we do not know how vulnerable
most systems are to warming of 1.5°, 2.0°, 2.5°C, etc.
The value 2.0°C is a reasonable target, but it is likely
chosen as some balance of safe and “achievable.” Some
estimates of the impacts of 2°C of warming include a
weakening of the thermohaline circulation, a higher
risk of a collapse of the Amazon, or a higher risk of the
loss of the Greenland ice sheet (Kriegler et al. 2009).
AMERICAN METEOROLOGICAL SOCIETY
However, the probability of these abrupt changes is
based largely on expert assessment (e.g., Kreigler et al.
2009; Smith et al. 2009), and it is very challenging to
quantify such expert assessment in terms of reliability
(Dawes et al. 1989).
There is some probability that 2°C of warming
is safe. We can estimate the emissions reductions
required to avoid 2°C or we can estimate the maximum permissible emissions (integrated over time)
that gives us a certain probability of avoiding 2°C
(Matthews et al. 2009; Meinshausen et al. 2009).
These depend on rates of emissions, rates of transitioning to a new non-carbon economy, climate
sensitivity, the probability of abrupt change, and
associated feedbacks in some key systems. Most
negotiations leading to Copenhagen were talking
about cuts in emissions of 10%, 20%, and 30% on 1990
levels, cuts per capita, or increasing the efficiency of
the carbon economy. None of these actually matter: it
is the actual concentration of CO2 in the atmosphere
that matters and how fast we can get the actual concentrations down to a level that we consider safe at
some level of agreed probability.
Achieving this goal is complex. The time scales
of emissions are important. Specifically, discussions
about emission reductions typically use different start
dates, different rates of emission reductions, or total
permissible emissions. There are arguments about
whether 2°C is safe. There are arguments about the
safe level of equilibrated CO2 , what CO2 can peak
at, and how fast it has to be reduced. Of course, all
arguments are complicated by questions of equity,
wrapped in rhetoric from some that there is not really
a problem and from others that there is an unmanageable catastrophe just around the corner. There is also
a position that cutting emissions threatens economic
growth and that we therefore need to maintain emissions to be able to afford to remedy the impacts of
emissions in the future.
All these positions may actually be consistent with
the best climate science, provided you are willing
to “cherry pick” that part of a PDF that suits your
belief. The IPCC Fourth Assessment Report provides
the PDFs of global warming in Fig. 6 of Alley et al.
(2007; the figure can be viewed online at www.ipcc.
ch/publications_and_data/ar4/wg1/en/spmsspmprojections-of.html). At the bottom of the PDF of
global warming by 2100, assuming a very low climate
sensitivity and very low emissions, is a value close
to 0°C. This would not be catastrophic, but it has a
very low probability (perhaps zero) of being achieved
both in terms of climate sensitivity and emissions. At
the very top end of the PDF with an approximately
AUGUST 2010
| 1009
equivalent probability is warming exceeding 8°C.
Rather than cherry pick particular points in this
distribution, communicators of the science need to
be very clear about the probabilities of the particular
future that they are predicting. Claims that Earth will
warm negligibly and claims that it will warm catastrophically could both be considered consistent with
the available science; however, those proposing these
two extreme positions should take care to emphasize
the near-zero probability of these outcomes.
Psychological phenomena relevant to comprehending the
problem and the solution. Table 3 highlights two relevant
psychological phenomena. The first is simply the
notion of a mental model and how it is used to organize
and evaluate information. As the previous section
illustrates, the mental model that is required to represent all of the variables and uncertainties involved in
assessing the threat of global warming is very complex.
By way of analogy, we might agree that the specific
details of the causal link between smoking and lung
cancer are complicated, but we are able to grasp that
inhaling carcinogenic material increases our chances
of lung cancer. For many of us, our fragmentary
knowledge of the links between cancer and smoking
is sufficient to represent the risks appropriately. In
contrast, understanding how and why an increase in
atmospheric CO2 leads to warming and how and what
we do (as individuals and communities) affects the
composition of the atmosphere is much harder.
Sterman and Booth Sweeney (2007; see also
Sterman 2008) provided some evidence on exactly
how difficult establishing the correct mental model
can be. They gave Massachusetts Institute of Technology (MIT) graduate students an excerpt of the
Summary for Policymakers from the IPCC Third
Assessment Report, which described the relationships among greenhouse gases, atmospheric concentrations, and global mean temperatures. They
then asked students first to estimate the future net
removal of CO2 from the atmosphere and then to
sketch on graphs the CO2 anthropogenic emissions
trajectory required to stabilize atmospheric CO2. In
84% of cases, participants drew emissions trajectories that violated basic principles of accumulation.
The trajectories reflected a mental model in which
atmospheric CO2 could be stabilized even if emissions
continuously exceeded removal. The authors provided an analogy that this is equivalent to assuming
that a bathtub continuously filled more quickly than
it drains will not overflow.
The difficulty in establishing the correct mental
model can be compounded by the second phenomenon listed in Table 3, the tendency for humans
to rely on confirmatory evidence. Many studies
have shown that when people test hypotheses they
tend to adopt a “positive” testing strategy, seeking
evidence that supports rather than challenges an
initial hypothesis. Although adopting such a testing
strategy is not necessarily bad practice (Klayman and
Table 3. Comprehending the solution: psychological factors that can influence the understanding of global
warming.
Psychological phenomenon
Illustration
Suggestion for tackling phenomenon
Construction of mental models
for representing problems
To facilitate understanding of a problem
(e.g., the threat of global warming), people
tend to assemble relevant but often
fragmentary prior knowledge and beliefs
into a “mental model.” This model is then
used to draw conclusions, assess risks, and
determine a course of action.
Try to gauge an audience’s mental model via
reactions to initial questions or presented
information and then tailor your message.
Establish basic concepts and lay the foundations for understanding the causes and
effects of global warming. Simple diagrams
and analogies (e.g., the bathtub; see text)
should aid this process.
Reliance on confirmatory
evidence
When testing hypotheses (e.g., that
human-induced global warming is a myth),
people often rely on “positive-testing
strategies,” asking questions that are expected to result in a “yes” response given
the truth of a working hypothesis. This
can lead to a “confirmation bias” if the
hypothesis is not properly evaluated.
Invite the audience to think about why a
working hypothesis might be incorrect.
Emphasize that all evidence, positive or negative, needs to be evaluated in the context of
the hypothesis [e.g., two consecutive years
that have shown cooling (positive evidence
for a myth hypothesis) in the context of
a warming trend over 50 yr (negative evidence)].
1010 |
AUGUST 2010
Ha 1987; McKenzie 2004), in some situations it can
result in inappropriately high confidence in a given
hypothesis, a phenomenon described as the “confirmation bias” (Wason 1960).
The relevance of confirmation bias to the construction of a mental model representing the risks of
global warming is clear; if people’s testing strategies
make it unlikely that they come across disconfirmatory evidence and they fail to take their strategy into
account when evaluating their hypothesis, the perception of the problem may be subsequently biased.
The tendency for people to downplay evidence that
is inconsistent with their working hypothesis or
interpret ambiguous evidence in ways that give their
hypothesis the benefit of the doubt (e.g., Klayman
1995) could also contribute to a failure to construct
appropriate mental models.
Tackling these phenomena entails patience, education, and careful framing. Misinformation, the occasional cold day, and the misquoting of science allow
a large fraction of the public and policy makers to
retain their existing mental models. Knowing your
audience and gauging its members’ mental models
is an important first step in tailoring presentation
of the science.
Consensus building: How group
dynamics can affect the formulation of the global warming
message. This penultimate section touches briefly
on factors affecting how the climate science community might improve the formulation of the message
before disseminating it to the wider community.
Fischhoff, writing presciently in 1981, noted that the
limitations of climate science (in terms of the certainties placed on predictions) could reach lay audiences
via explicit statements of uncertainties, which entail
the types of problems discussed in the preceding
sections, and the observation of disagreement among
experts. In the latter case, Fischhoff (1981, p. 178)
worried about the following:
Unless the audience has an appreciation of the naturally disputative and accretive character of science,
its resolution of the conflict may not be a balanced
and informed weighting of the sides. Alternative
resolutions are doubting the probity of the disputants, siding with the most assertive (or colorful or
optimistic or certain), or deciding that “anything
goes” and that “my guess is as good as yours.”
The irony of the situation almost 30 yr later is that
there is little genuine disagreement among experts
AMERICAN METEOROLOGICAL SOCIETY
about the basic science of global warming, but the
desire for “balanced” media coverage of the debate
appears to have led to exactly the outcome Fischhoff
feared (Moser and Dilling 2004): A recent poll found
that 40% of surveyed Americans believe that there
is “a lot of disagreement” among scientists about
whether global warming is happening (Leiserowitz
et al. 2010).
This outcome is deeply confronting to the climate science community; the climate scientists use
the scientific method to reject wrong science. Once
wrong science has been rejected, the debate moves
on. Unfortunately the scientific method is a scientific
method and, as Fischhoff noted, it is not intuitive to
many members of the general public or politicians
who are more familiar with arguments won by clever
debate or oratory. How to maintain the scientific
method and convey the urgency of the message without resorting to public advocacy is a tricky path to
follow (cf. Fischhoff 2007).
One route to remedying this problem is ensuring
that climate scientists emphasize the transparent
and unbiased nature of their discussions and evaluation of evidence. A starting point is to ensure that,
in meetings where consensus between experts is the
goal, steps are taken to encourage optimal discussion
of information.
Climate community phenomenon: The experts in a room
problem. The IPCC or equivalent assessments of the
state of the science of climate and global warming
are phenomenally rigorous. A key part of the process
is lead author meetings where around 10 experts
(lead authors) meet to discuss specific issues, specific reviewers’ comments, or contentious pieces of
science. A consensus is reached, and this consensus,
provided it can be fully supported by the literature,
is reported.
One issue that might contribute to problems in
reaching a consensus is bias in the literature we read.
Very high impact journals such as Nature and Science
publish the latest high impact science. If experts
source their perspective from journals such as Nature
and Science they could form a different opinion than
if they sourced their perspective from the Journal of
Climate.
Psychology factors affecting the process and perception
of consensus building. Table 4 presents two psychological phenomena relevant to consensus building
and group dynamics. The first, biased information
pooling, refers to a process that can occur within
a group attempting to reach a consensus. Pooling
AUGUST 2010
| 1011
Table 4. Consensus building: psychological factors that can influence the process and public perception of
how a consensus about climate science is reached.
Psychological phenomenon
Illustration
Suggestion for improving process
Biased information pooling in
group discussion
The effective pooling of information in
group discussions can fail to occur because
discussion is dominated by information that
is shared by members prior to discussion
and by information that is consistent with
members’ prior beliefs and preferences.
Allowing sufficient discussion time for
“unshared” information to be aired
can counteract biased pooling, as can
having at least one group member who
is an advocate of the position favored by
unshared information.
Groupthink
A label to describe defective decision making
that can arise from highly cohesive, insular
groups that have directed leadership, a lack
of procedures for search and appraisal of
information, and low confidence in the ability to find an alternative solution to the one
favored by the leader.
To avoid public perception that the climate science community has been beset
by “groupthink,” steer clear of an us
(believers) vs them (deniers) portrayal
in your presentations of the scientific
evidence.
information should lead to improved decisions,
provided information is shared effectively. However,
research indicates that effective pooling often fails
to occur, because group discussion is dominated by
information that members hold in common prior
to discussion and by information that is consistent
with members’ prior beliefs and preferences (Stasser
and Titus 1985; see also the positive testing strategy
described earlier). This can have the deleterious effect
of perpetuating rather than correcting preexisting
biases. Thus, if experts’ initial perspectives are biased
in idiosyncratic ways (perhaps because of the overweighting of information from high impact sources),
these biases will not necessarily be eroded through
discussion.
The second phenomenon in Table 4 is “groupthink,”
a term coined by Janis (1972) that has become a catchall label for defective decision making that can arise
from groups with dysfunctional dynamics (Fuller and
Aldag 1998). We are not asserting that groupthink has
taken hold of the climate science community; this is
highly unlikely because there is too much to be gained
for individuals who provide rigorous and scientifically defensible alternative ideas. However, the public
perception and media portrayal of the community
as being unwilling to listen to dissenters combined
with an unassailable belief in the correctness of their
position (antecedent conditions of groupthink) need
to be addressed.
Table 4 notes two brief suggestions for improving
the process and perception of consensus building.
On the process side, other techniques for improving
effective sharing of information in discussion include
ensuring access to all relevant informational records
1012 |
AUGUST 2010
during discussion, separating the process into a
search for information followed by the integration
of that information, and assigning group members to
be responsible for specific categories of information
and making sure that knowledge of who knows what
is shared among group members (Kerr and Tindale
2004). On the perception side, climate scientists must
strive to be trusted, credible sources of information
without, if possible, engaging in advocacy and divisive
rhetoric. In short, try to let the science speak for itself
in a clear, comprehensible manner (Fischhoff 2007).
Concluding remarks and summary.
Simply presenting the facts and figures about global
warming has failed to convince large portions of the
general public, journalists, and policy makers about
the scale of the problem and the urgency of required
action. From a psychologist’s perspective this disconnect is not surprising. Facts and figures need
to be tailored to fit with the way in which humans’
process information, deal with uncertainty, and form
attitudes and opinions. By combining our efforts,
climate scientists and psychologists can provide not
just the science but the tools for communicating and
interpreting the science (Sterman 2008).
We are not suggesting that psychology has all the
answers when it comes to convincing the public about
the need for action. Stern (1992) pointed out that
many variables outside the scope of psychological
theory (e.g., socio-demographic status, geographic
context, institutional arrangements) will all have a
significant impact on the willingness to engage with
the science and commit to action. Global warming
is also an incredibly politicized issue; thus, many of
the pronouncements on the problem and the solution
are subjected to considerable political “spin” that
goes beyond the scientific domain in which climate
scientists feel comfortable. Psychology can provide
some insight into how these issues can be addressed,
but other branches of social science such as sociology
and social policy also have much to offer (Fischhoff
2007; Stern 1992).
In summary, our review highlights four classes
of psychological phenomena that provide food for
thought to the climate scientist wishing to disseminate findings to the wider community: 1) sampling
issues: clarity about the source and representativeness
of samples of evidence that your audience and you
are using to form inferences and draw conclusions;
2) framing issues: methods for presenting science
should engage cognitive and emotional processing,
in a balanced manner, and try to make distant future
outcomes concrete; 3) comprehending the problem
and solution: communicators should take into
account the “mental model” held by members of their
audience and tailor presentations accordingly; and 4)
consensus building: the process and public perception
of reaching a consensus about the science needs to be
effective, transparent, and objective.
As we noted at the outset, our treatment of these
issues has been brief and our suggestions simple;
many of the papers we have cited contain more
detailed empirical demonstrations of the phenomena along with specific strategies and protocols for
addressing them (see Further Reading for more
information). Our hope is that this overview of the
essence of what psychology can offer will precipitate
further much needed collaboration between our two
communities and ultimately lead to the message of
global warming being heard and heeded.
References
Alley, R. B., and Coauthors, 2007: Summary for policymakers. Climate Change 2007: The Physical Science
Basis, S. Solomon et al., Eds., Cambridge University
Press, 1–18.
Budescu, D. V., S. Broomell, and H. Por, 2009: Improving
communications of uncertainty in the reports of
the Intergovernmental Panel on Climate Change.
Psychol. Sci., 20, 299–308.
Dawes, R. M., D. Faust, and P. E. Meehl, 1989: Clinical
versus actuarial judgment. Science, 243, 1668–
1674.
Easterling, D. R., and M. F. Wehner, 2009: Is the climate
warming or cooling? Geophys. Res. Lett., 36, L08706,
doi:10.1029/2009GL037810.
AMERICAN METEOROLOGICAL SOCIETY
Fiedler, K., 2000: Beware of samples! A cognitiveecological sampling approach to judgment biases.
Psychol. Rev., 107, 659–676.
Fischhoff, B., 1981: Hot air: The psychology of COinduced climatic change. Cognition, Social Behavior
and the Environment, J. Harvey, Ed., Erlbaum,
163–184.
—, 2007: Non-persuasive communication about
matters of greatest urgency: Climate change. Environ.
Sci. Technol., 41, 7204–7208.
—, and L. Furby, 1983: Psychological dimensions of
climatic change. Social Science Research and Climate
Change, R. S. Chen, E. Boulding, and S. H. Schneider,
Eds., D. Reidel, 183–203.
Fuller, S., and R. Aldag, 1998: Organizational Tonypandy:
Lessons from a quarter century of the groupthink
phenomenon. Organ. Behav. Hum. Decis. Processes,
73, 163–184.
Gigerenzer, G., and U. Hoffrage, 1995: How to improve
Bayesian reasoning without instruction: Frequency
formats. Psychol. Rev., 102, 684–704.
Hardisty, D. J., and E. U. Weber, 2009: Discounting
future green: Money versus the environment. J. Exp.
Psychol., 138, 329–340.
Janis, I., 1972: Victims of Groupthink: A Psychological
Study of Foreign-Policy Decisions and Fiascoes.
Houghton Mifflin, 277 pp.
Kahneman, D., and S. Frederick, 2002: Representativeness revisited: Attribute substitution in intuitive
judgment. Heuristics and Biases, T. D. Gilovich,
D. W. Griffin, and D. Kahneman, Eds., Cambridge
University Press, 49–81.
Kerr, N. L., and R. S. Tindale, 2004: Group performance and decision making. Ann. Rev. Psychol., 55,
623–655.
Klayman, J., 1995: Varieties of confirmation bias.
Psychol. Learn. Motiv., 32, 385–418.
—, and Y.-W. Ha, 1987: Confirmation, disconfirmation,
and information in hypothesis testing. Psychol. Rev.,
94, 211–228.
Kriegler, E., J. W. Hall, H. Held, R. Dawson, and H.-J.
Schellnhuber, 2009: Imprecise probability assessment
of tipping points in the climate system. Proc. Nat.
Acad. Sci., 106, 5041–5046.
Larrick, R. P., 2004: Debiasing. The Blackwell Handbook
of Judgment and Decision Making, D. J. Koehler and
N. Harvey, Eds., Blackwell, 316–337.
Leiserowitz, A., 2006: Climate change risk perception
and policy preferences: The role of affect imagery
and values. Climatic Change, 77, 45–72.
—, E. Maibach, and C. Roser-Renouf, 2010: Climate
change in the American Mind: Americans’ global
warming beliefs and attitudes in January 2010. Yale
AUGUST 2010
| 1013
University and George Mason University Project on
Climate Change, 10 pp. [Available online at http://
environment.yale.edu/uploads/AmericansGlobalW
armingBeliefs2010.pdf.]
Linville, P. W., and G. W. Fischer, 1991: Preferences for
separating and combining events: A social application of prospect theory and the mental accounting
model. J. Pers. Soc. Psychol., 60, 5–23.
Loewenstein, G. F., E. U. Weber, C. K. Hsee, and E. S.
Welch, 2001: Risk as feelings. Psychol. Bull., 127,
267–286.
Mariconti, C., 2009: Understanding the disconnect on
global warming. Observer, 22, 17–19.
Matthews, H. D., N. P. Gillett, P. A. Stott, and
K. Zickfeld, 2009: The proportionality of global
warming to cumulative carbon emissions. Nature,
459, 829–832.
McKenzie, C. R. M., 2004: Hypothesis testing and
evaluation. The Blackwell Handbook of Judgment and
Decision Making, D. J. Koehler and N. Harvey, Eds.,
Blackwell, 200–219.
Meinshausen, M., N. Meinshausen, W. Hare, S. C. B.
Raper, K. Frieler, R. Knutti, D. J. Frame, and M. R.
Allen, 2009: Greenhouse-gas emission targets for
limiting global warming to 2°C. Nature, 458,
1158–1163.
Morgan, M. G., B. Fischhoff, A. Bostrom, and C. J.
Atman, 2002: Risk Communication: A Mental Models
Approach. Cambridge University Press, 351 pp.
Moser, S. C., and L. Dilling, 2004: Making climate hot:
Communicating the urgency and challenge of global
climate change. Environment, 10, 32–46.
Newell, B. R., C. J. Mitchell, and B. K. Hayes, 2008:
Getting scarred and winning lotteries: Effects of
exemplar cuing and statistical format on imagining
low-probability events. J. Behav. Decis. Making, 21,
317–335.
Nicholls, N., 1999: Cognitive illusions, heuristics, and
climate prediction. Bull. Amer. Meteor. Soc., 80,
1385–1397.
Pachauri, R. K., and A. Reisinger, Eds., 2007: Climate
Change 2007: Synthesis Report. Cambridge University
Press, 104 pp.
Slovic, P., J. Monahan, and D. M. MacGregor, 2000:
Violence risk assessment and risk communication:
The ef fects of using actua l cases, prov iding
instructions, and employing probability vs. frequency
formats. Law Hum. Behav., 24, 271–296.
1014 |
AUGUST 2010
Smith, J. B. and Coauthors, 2009: Assessing dangerous
climate change through an update of the Intergovernmental Panel on Climate Change (IPCC)
“reasons for concern.” Proc. Nat. Acad. Sci., 106,
4133–4137.
Somers, J., 2009: AAAS joins leading scientif ic
organizations in letter to senators reaffirming
scientific consensus on climate change. American
Association for the Advancement of Science News
Release. [Available online at www.aaas.org/news/
releases/2009/1021climate_letter.shtml.]
Stasser, G., and W. Titus, 1985: Pooling of unshared
information in group decision making: Biased
information sampling during discussion. J. Pers. Soc.
Psychol., 48, 1467–1478.
Sterman, J. D., 2008: Risk communication on climate:
Mental models and mass balance. Science, 322,
532–533.
—, and L. Booth Sweeney, 2007: Understanding public
complacency about climate change: Adults’ mental
models of climate change violate conservation of
matter. Climatic Change, 80, 213–238
Stern, P. C., 1992: Psychological dimensions of global
environmental change. Annu. Rev. Psychol., 43,
269–302.
Strack, F., and T. Mussweiler, 1997: Explaining the
enigmatic anchoring effect: Mechanisms of selective
accessibility. J. Pers. Soc. Psychol., 73, 437–446.
Trope, Y., and N. Liberman, 2003: Temporal construal.
Psychol. Rev., 110, 403–421.
Tversky, A., and D. Kahneman, 1974: Judgment under
uncertainty: Heuristics and biases. Science, 185,
1124–1131.
—, and —, 1981: The framing of decisions and the
psychology of choice. Science, 211, 453–458.
—, and —, 1992: Advances in prospect theory:
Cumulative representation of uncertainty. J. Risk
Uncertainty, 5, 297–323.
Wason, P. C., 1960: On the failure to eliminate hypotheses
in a conceptual task. Quart. J. Exp. Psychol., 12,
129–140.
Weber, E. U., 2006: Experience-based and descriptionbased perceptions of long-term risk: Why global
warming does not scare us (yet). Climatic Change,
77, 103–120.
Yamagishi, K., 1997: When a 12.86% mortality is
more dangerous than 24.14%: Implications for risk
communication. Appl. Cognit. Psychol., 11, 495–506.